Application of the Grade of Membership model (GoM) to delineate an Urban Hierarchy in Brazilian Amazonia

Douglas Sathler 55-31-34270841 douglas@cedeplar.ufmg.br Geographer, Ph.D. student, Graduate Program in Demography, Center for Regional Development and Planning (Cedeplar/UFMG) Belo Horizonte/Minas Gerais/Brazil

Roberto L. Monte-Mór 55-31-3482-8007 monte-mor@cedeplar.ufmg.br Architect, Ph.D. in Urban Planning (UCLA), Associate Professor at Cedeplar/UFMG, Center for Regional Development and Planning, Federal University of Minas Gerais Belo Horizonte/Minas Gerais/Brazil

Carla Jorge Machado 55-31-3409-7159 carla@cedeplar.ufmg.br Economist, Ph.D. in Population dynamics (Johns Hopkins University), Associate Professor at Cedeplar/UFMG, Center for Regional Development and Planning, Federal University of Minas Gerais Belo Horizonte/Minas Gerais/Brazil

Alfredo Costa 55-31-3581-1016 alfredo.ufmg@yahoo.com.br Undergraduate Student in Geography at IGC/UFMG, Institute of Geosciences, Federal University of Minas Gerais Belo Horizonte/Minas Gerais/Brazil

239 (around 72%) municipalities in the rest of the region presented population smaller than 20 thousand people. “In the last decades. the recent birth of urban centers with impressive population in the inner region has assured more balance to the Amazon urban net. and it continued to decrease even more between 1991-2000 (2. the rising of middle-size cities and small towns seem to break the huge difficult to populate a significant part of Amazonia. Notwithstanding the significant demographical growth of the state capitals in the past decades. in all parts of the country. However. including 12 municipalities with more than 100 thousand people and 31 municipalities with population between 50 and 100 thousand residents. by an explosion of little municipalities as a result of the emancipations that occurred in that period of time. spread around the forest and connected only through streams. there were a little more than 500 municipalities in Amazonia and. The decade of 1990 was marked.51% per year. among those of them. the researchers cold verify a true urban explosion near to the main regional roads. most of the new territorialities were born with less than 20 thousand residents.43 % per year) had been higher than the national average rates in that same period of time (2. In that period. Table 1 presents the municipalities distribution for population size groups between 1970 and 2007. despite the growth of 140 new territorial unities because of ongoing emancipation processes in the last decade. In addition to this. the participation of the municipalities with middle or large population size kept rising. Amazonia was known for its small number of cities and towns. Three decades ago. Belém and Manaus exceeded 1 million people and. there were only twenty municipalities with population larger than 50 thousand inhabitants.5 % per year). 255 new municipalities in Amazonia arose.48 % per year). Between 1980 and 1991 the GGR value in that region decreased to 3. As such. In other words. the recently observed GGRs in Amazonia were still in relatively higher levels when compared to the other Brazilian regional values. with population of up to 20 thousand people. In that context. in that year. the Geometric Growth Rate (GGR) for the Legal Amazonia population between 1970 and 1980 (4. Farther on.Extended Summary The Amazonian urban network has been developing along with the arisal of middle size cities and the multiplication of new towns and small urban agglomerations along the main regional roads and rivers. 2005). Furthermore. Those expressive decreases of the GGR occurred because of reorientation of the internal migration movements within the country. In 1970. had been rising. Considering the recent history of the cities` growth. the complexity of the Amazonian urban network increased in 1980. The growth rates were higher than the national averages due those migration flows. In fact. the percentage participation of that group of cities. the municipalities with more than 50 thousand residents represented around 12% of all urban centers in the region. which was the only way to integrate the localities. in one of the largest migration movements ever” (Matos. reaching values near to what can be observed in 1970. among which only five had more than 100 thousand inhabitants. we . especially those that came from to the Northwest and South of Brazil. Most of the population was engaged in very different activities as compared to those in urban centers and the regional economy was essentially focused on the occupations related to rivers and forests. According to the Brazilian Demographic Census Bureau. 16 had a population between 100 thousand and 1 million residents. In 1991. In that year.

The number of municipalities with more than 100.16 0.000 increased to 49 in the last analyzed year.00 100 1991 303 142 43 16 2 506 (%) 59. the present study aims to contribute for a better comprehension of the urban hierarchy and for the relationship between the cities in Amazonia.15 0.88 28.000.000 and 100.65 2.000 more than 1.000 50.000 to 50.26 100 According to the most recent dataset. There are several areas in which the urban tissue expansion was more intense. Southeast and East of Amazonia and following the margin of Amazonas River to Manaus. Table 1 .0000 100.99 21. .have to highlight that the number of municipalities with more than 100 thousand people increased from 18 to 21 between 1991 and 2000.34 6.38 27. for instance near the main roads that cross the South.000 to 1.50 0. that model permits describing the differences between these city groups. 1970 to 2000.94 5. 1970 239 73 15 5 0 332 (%) 71. the South of Roraima and the North of Amapá. Moreover. it seems clear that the modifications between 2000 and 2007 were concentrated in the middle-sized cities in Amazonia.87 8. Map 1 clearly shows the impact of the main roads in the development of urban agglomerations in Amazonia.50 3. However. considerable parts of Amazonas state.52 1. the municipalities with population between 50.000 Total Source: IBGE.28 0. In such a situation. In that context.99 4.65 21. In addition.000 to 10. Likewise.51 0.000 20. we can highlight the Southwest and the Northwest of Pará.00 100 1980 221 102 31 12 0 366 (%) 60.Municipalities in Amazonia sorted by population size groups between 1970 e 2007 Inhabitants number less than 20.000.47 3. it is suggested the application of the Grade of Membership model (GoM) that is useful to delineate profiles and to identify city groups with similar characteristics.40 100 2000 530 167 43 19 2 761 (%) 69.81 22. against 43 in 2000.06 8. For this task. there are huge areas without occupation according to that map.26 100 2007 516 170 49 24 2 761 (%) 67.000 habitants was 19 in 2000 and it increased to 24 in 2007.44 3.

4 .Infrastructure e services 5 . in general.1 .% of people with computer 5. In GoM.1 .4 .Number of registrations in college 4.1 .12 . In that sense.6 .Municipality geometric growth rate between 2000 and 2007 2.Goods acecibility 6 . the variables which measure the functionality and the basic and specialized service offers are very useful.Number of registrations in elementary school 4.Socioeconomic 4 .Differently than most of cluster analysis statistic methods. the same observation could have a determined grade of membership in multiples groups and. It is valid to assume the idea that.Variable that represents the number of times that na analyzed city was mentioned as the nearest urban center with larger population 2. the model intends to consider aspects beyond the pure economic and demographic analysis.6 .3 .Poverty proportion 3.Number of high schools 4.11 . For this purpose.3 .9 .3 . because of this propriety.5 -% of people with phone 6.3 . GoM also has the capacity to analyze categorical data in short samples with a large number of variables.Outpatient units 4.4 . the greater`s regional .Number of elementary schools 4.13 .1 .4 . Likewise.Municipality participation fund 3.spatial Table 2 .Gross Domestic Product (GDP) 3. In this paper.1 .Health station 4.2 . the greater distance between a certain city and a more populated center.Functionality The diverse sorts of variables present in the model aim to establish a classification of the greatness and influence capacity of the cities in Amazonia.% of people with eletric energy at home 4.% of people with TV 5.1 .% of people with access to the garbage collecting 4.2 .5 .Internal variables to the GoM model Variáveis 1.2 . the indexes which measure the goods access and the other that measure equipments and infra-structures present in the cities.HDI (Human Development Index) 4. according to the table 2.Hospital beds 4.Fleet of vehicles 5.Municipality urbanization degree 2. Natureza das variáveis 1 .Municipality population in 2007 2. In addition.A variable will be created that will represent the number of funcionalities of the municipality to a total of functions that will be predefined Source: Elaborated by Douglas Sathler 2 . it is called fuzzy model.% of people with access to treated water 4.Health centers 5.Demographic 3 .10 .% of people with refrigerator 5.Number of registrations in high school 4.Real distance between na analyzed city and a more populated center 1.Regional geometric growth rate between 2000 and 2007 3.6 .2 .Hospital 4.2 . the model will be helpful to select and identify the hierarchy patterns between the cities in Amazonia with the support of variables of diverse natures. The incorporation of spatial variables seems to be very important because the urban hierarchy is also defined by the influence of the city net distribution. GoM do not consider that people and objects are organized in well defined groups. with clear impacts on the ways they relate.

Understanding the hierarchy organization of the cities in Amazonia seems to be an interesting exercise for the comprehension of the regional net urban dynamics. Alexandrino. D.1. Tese (Thesis) – Center for Regional Development and Planning (Cedeplar/UFMG). . R. In that sense. A. R. Ciência e Saúde Coletiva (Journal). especially when it comes to the access to services of several types and levels of sophistication. References Cerqueira. C. Likewise. 2005. I. Matos. 7(4): 757 – 76.importance. Espacialidades em rede: População. urbanização e migração no Brasil contemporâneo. Perfis de utilização de serviços de saúde no Brasil. R. as can be seen in variable 1. A more equal population distribution in the Amazon territory could bring a series of benefits. the necessity of policies that try to establish more structured urban nets in Amazonia seems evident. Leite. 2000. T. Belo Horizonte. a certain city takes on a leading role if mentioned often as being the nearest urban center with a large population. Tipologia e características dos estabelecimentos escolares brasileiros. Sawyer. 2004.. O. C.